What's with the 2sk3018 MOSFET?

[Winfield Hill]

The 2sk3018 MOSFET is an unusual animal.
It holds the top 30V spot, in my giant
MOSFSET table. Its Crss = 4pF at 5V is
the lowest, and its Ron(max) = 8 ohms,
with 4V drive, is the highest, giving it
the top spot. It's likely the ideal part
for my fast programmable current source.

But oops, Rohm says it's NRND. However,
it's 2nd sourced by MCC, and Invac and
HTsemi. And Kexin, distributed by LCSC.

Whoa, five different sources! Is / was
this a jellybean? What are / were its
big applications? Anybody hear of it?


--
Thanks,
- Win

 

[Winfield Hill]

Winfield Hill wrote...

Winfield Hill wrote...

The 2sk3018 MOSFET is an unusual animal.
It holds the top 30V spot, in my giant
MOSFET table. Its Crss = 4pF at 5V is
the lowest, and its Ron(max) = 8 ohms,
with 4V drive, is the highest, giving it
the top spot. It's likely the ideal part
for my fast programmable current source,

which will run cascode into a HV MOSFET.

But oops, Rohm says it's NRND. However,
it's 2nd sourced by MCC, and Invac and
HTsemi. And Kexin, distributed by LCSC.

Whoa, five different sources! Is / was
this a jellybean? What are / were its
big applications? Anybody hear of it?

More claims to fame: specified for 2.5V
gate drive, small SC-70 package, $0.05
from Future. And a sixth manufacturer!!,
Changjiang, also stocked and sold by LCSC.

Price, 1.5 cents.


--
Thanks,
- Win

 

[Winfield Hill]

Winfield Hill wrote...

The 2sk3018 MOSFET is an unusual animal.
It holds the top 30V spot, in my giant
MOSFET table. Its Crss = 4pF at 5V is
the lowest, and its Ron(max) = 8 ohms,
with 4V drive, is the highest, giving it
the top spot. It's likely the ideal part
for my fast programmable current source,

which will run cascode into a HV MOSFET.

But oops, Rohm says it's NRND. However,
it's 2nd sourced by MCC, and Invac and
HTsemi. And Kexin, distributed by LCSC.

Whoa, five different sources! Is / was
this a jellybean? What are / were its
big applications? Anybody hear of it?

More claims to fame: specified for 2.5V
gate drive, small SC-70 package, $0.05
from Future. And a sixth manufacturer!!,
Changjiang, also stocked and sold by LCSC.


--
Thanks,
- Win

 

[John Larkin]

On 4 Aug 2019 16:05:31 -0700, Winfield Hill <winfieldhill@yahoo.com>
wrote:

The 2sk3018 MOSFET is an unusual animal.
It holds the top 30V spot, in my giant
MOSFSET table. Its Crss = 4pF at 5V is
the lowest, and its Ron(max) = 8 ohms,
with 4V drive, is the highest, giving it
the top spot. It's likely the ideal part
for my fast programmable current source.

But oops, Rohm says it's NRND. However,
it's 2nd sourced by MCC, and Invac and
HTsemi. And Kexin, distributed by LCSC.

Whoa, five different sources! Is / was
this a jellybean? What are / were its
big applications? Anybody hear of it?

HV cascodes can have some interesting hazards when they turn off fast.
You can blow out the gate of the top fet, or the drain of the little
one below. An analog current controller isn't so scary as a fast
switch.

We considered a GaN+SiC cascode for fast HV pulsing but decided it was
better to just build a monster gate driver for the SiC.



--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[Winfield Hill]

John Larkin wrote...

HV cascodes can have some interesting hazards
when they turn off fast. You can blow out
the gate of the top fet, or the drain of the
little one below.

Sure, but easily handled with a gate-source
TVS, e.g. see Figures 3x.56 and 4x.147 in
my x-Chapter handouts.

I haven't seen trouble with cascode MOSFETs,
even fast switched ones, but worry about HV
supply being suddenly shorted, or applied.

I've been using Diodes' D1213A-01, with a
5A 8/20μs pulse rating, 1.2 pF, in SOD323
package. Happy to spend more capacitance
on the TVS, for a higher current rating,
but haven't found one.


--
Thanks,
- Win

 

[Robert Baer]

Winfield Hill wrote:

The 2sk3018 MOSFET is an unusual animal.
It holds the top 30V spot, in my giant
MOSFSET table. Its Crss = 4pF at 5V is
the lowest, and its Ron(max) = 8 ohms,
with 4V drive, is the highest, giving it
the top spot. It's likely the ideal part
for my fast programmable current source.

But oops, Rohm says it's NRND. However,
it's 2nd sourced by MCC, and Invac and
HTsemi. And Kexin, distributed by LCSC.

Whoa, five different sources! Is / was
this a jellybean? What are / were its
big applications? Anybody hear of it?

WARNING. Be advised that every manufacturer of a given part (the
2SK3018 in this case) runs their fab differently, and you may find that
those particular specs may be WIDELY different - enough to make the part
virtually useless.

 

On Tuesday, 6 August 2019 07:00:09 UTC+1, Robert Baer wrote:

Winfield Hill wrote:
The 2sk3018 MOSFET is an unusual animal.
It holds the top 30V spot, in my giant
MOSFSET table. Its Crss = 4pF at 5V is
the lowest, and its Ron(max) = 8 ohms,
with 4V drive, is the highest, giving it
the top spot. It's likely the ideal part
for my fast programmable current source.

But oops, Rohm says it's NRND. However,
it's 2nd sourced by MCC, and Invac and
HTsemi. And Kexin, distributed by LCSC.

Whoa, five different sources! Is / was
this a jellybean? What are / were its
big applications? Anybody hear of it?


WARNING. Be advised that every manufacturer of a given part (the
2SK3018 in this case) runs their fab differently, and you may find that
those particular specs may be WIDELY different - enough to make the part
virtually useless.

I thought the whole point of part numbers was that any device that met its specs could get the part number. Other specs may differ certainly, but not those defining ones.


NT

 

[Winfield Hill]

Robert Baer wrote...

Winfield Hill wrote:
The 2sk3018 MOSFET is an unusual animal.
It holds the top 30V spot, in my giant
MOSFSET table. Its Crss = 4pF at 5V is
the lowest, and its Ron(max) = 8 ohms,
with 4V drive, is the highest, giving it
the top spot. It's likely the ideal part
for my fast programmable current source.

But oops, Rohm says it's NRND. However,
it's 2nd sourced by MCC, and Invac and
HTsemi. And Kexin, distributed by LCSC.

Whoa, five different sources! Is / was
this a jellybean? What are / were its
big applications? Anybody hear of it?

WARNING. Be advised that every manufacturer of a given part (the
2SK3018 in this case) runs their fab differently, and you may find that
those particular specs may be WIDELY different - enough to make the part
virtually useless.

Or not. Of course, one must fully evaluate a
second-source manufacturer's parts, before
committing it to manufacturing.

But in the specific case of the 2sk3018, there
are several distinguishing features that most
of the 2nd-source providers will probably meet:

(1) The high 4-volt Ron spec. This insures a
small die, which a manufacturer will cling too,
to insure higher profits. It's most important
to me, because it goes with very low capacitance.
Capacitance is the primary parameter I'll check.

(2) The 2.5-volt logic-drive capability. Along
with low cost, and the tiny SC70 package, this
is probably the spec that's most responsible for
the popularity of this part. It's not important
to me, but may help to insure interchangeability.

The very existance of these alternate sources,
each with active stocking distributors, helps
to ameliorate Rohm's NRND. I've also ordered
100 of the MCC parts, to evaluate with Rohm.
Maybe I'll also order strips of Changjiang and
Kexin parts from LCSC, they're dirt cheap. A
full 3000-piece reel of Changjiang's part is $40.


--
Thanks,
- Win

 

[John Larkin]

On 6 Aug 2019 02:24:16 -0700, Winfield Hill <winfieldhill@yahoo.com>
wrote:

Robert Baer wrote...

Winfield Hill wrote:
The 2sk3018 MOSFET is an unusual animal.
It holds the top 30V spot, in my giant
MOSFSET table. Its Crss = 4pF at 5V is
the lowest, and its Ron(max) = 8 ohms,
with 4V drive, is the highest, giving it
the top spot. It's likely the ideal part
for my fast programmable current source.

But oops, Rohm says it's NRND. However,
it's 2nd sourced by MCC, and Invac and
HTsemi. And Kexin, distributed by LCSC.

Whoa, five different sources! Is / was
this a jellybean? What are / were its
big applications? Anybody hear of it?

WARNING. Be advised that every manufacturer of a given part (the
2SK3018 in this case) runs their fab differently, and you may find that
those particular specs may be WIDELY different - enough to make the part
virtually useless.

Or not. Of course, one must fully evaluate a
second-source manufacturer's parts, before
committing it to manufacturing.

But in the specific case of the 2sk3018, there
are several distinguishing features that most
of the 2nd-source providers will probably meet:

(1) The high 4-volt Ron spec. This insures a
small die, which a manufacturer will cling too,
to insure higher profits. It's most important
to me, because it goes with very low capacitance.
Capacitance is the primary parameter I'll check.

(2) The 2.5-volt logic-drive capability. Along
with low cost, and the tiny SC70 package, this
is probably the spec that's most responsible for
the popularity of this part. It's not important
to me, but may help to insure interchangeability.

The very existance of these alternate sources,
each with active stocking distributors, helps
to ameliorate Rohm's NRND. I've also ordered
100 of the MCC parts, to evaluate with Rohm.
Maybe I'll also order strips of Changjiang and
Kexin parts from LCSC, they're dirt cheap. A
full 3000-piece reel of Changjiang's part is $40.

I've found wide variation in switching speed on 2N7002s and BSS123s.
We sole-source spec the Fairchild parts. And hope they don't change.




--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[Winfield Hill]

John Larkin wrote...

I've found wide variation in switching speed
on 2N7002s and BSS123s. We sole-source spec
the Fairchild parts. And hope they don't change.

But now that Fairchild is part of ON Semi, and
ON Semi has conglomerated pretty much everything,
how can you be sure that you won't get the
ON Semi version of the 2N7002? You should order
a lifetime supply. They're not very expensive.


--
Thanks,
- Win